Research On Intelligent Scheduling Of Grinding Process In Multi Variety Single Piece And Small Batch Sand Casting Enterprises

Grinding process scheduling is a decision-making process of making the optimal grinding task allocation scheme in multi variety single piece and small batch sand casting enterprises under complex constraints such as different skill levels of workers and different grinding requirements of castings.It is a key link in production planning of casting enterprises.Grinding process scheduling can be abstracted as a parallel machine scheduling problem.All grinding workers constitute a parallel and independent processing environment.The parallel machine scheduling problem under complex constraints belongs to NP hard problem,and it is difficult to obtain the optimal solution in acceptable time by accurate algorithm.In view of this problem,the existing manual production scheduling mode has some prominent problems,such as time-consuming and laborious,unbalanced task allocation,mismatch between task requirements and workers' skill level,unstable production scheduling effect and so on,which lead to unreasonable allocation of enterprise resources,low production efficiency and casting accumulation in some workers.Therefore,this paper proposes an intelligent algorithm aided production scheduling model,studies the mathematical modeling and solving algorithm of parallel machine scheduling of grinding process in Foundry Enterprises Based on improved genetic algorithm,and puts forward several improvement strategies for standard genetic algorithm,and verifies the excellent performance of the proposed improved genetic algorithm and intelligent algorithm through the comparison of different scale simulation experiments and actual production scheduling results The significant advantages of the model of assistant scheduling.The main contents of this paper are as followsFirstly,the intelligent scheduling model of grinding process is constructed.The first step is to establish a grinding wage coefficient calculation scheme,and determine a grinding wage coefficient according to the casting type,weight,material,surface roughness requirements and pickling requirements,so as to quantify the grinding workload of a single casting and provide the basis for the balance of grinding task allocation.The second step is to use natural language reasoning mechanism and triangular fuzzy number to determine the grinding time.The third step is to build an integer programming mathematical model with the optimization objectives of minimizing the difference of the sum of grinding wage coefficients,minimizing the maximum completion time,and minimizing the mismatch between workers' skill level and grinding task requirements,so as to provide the model basis for solving grinding process scheduling plan.Secondly,a parallel machine scheduling solution of grinding process based on improved genetic algorithm is proposed.In the first step,according to the requirement level of casting surface roughness and the skill level of grinding workers,the casting and workers are divided and combined to generate different scheduling task subsets.In the second step,an improved genetic algorithm is proposed,which is used to solve each task data set and get the optimal grinding task scheduling scheme.The improvement strategies of genetic algorithm mainly include: using heuristic decoding rules to optimize the quality of the initial solution;designing elite selection operator to improve the global search ability;using Hamming crossover operator to prevent close relatives from intersecting;embedding tabu search algorithm in mutation operator to increase the local search ability,etc.Finally,the algorithm improvement strategy comparison test,multiple algorithm performance comparison test and actual scheduling effect verification are carried out.The first step is to combine the standard genetic algorithm with different improvement strategies to form multiple genetic algorithm variants,and compare and test the performance indexes of each variant in solving different scale simulation experiments;the second step is to compare and test the performance indexes of improved genetic algorithm and other four typical algorithms in solving different scale simulation experiments;the third step is to compare and analyze the performance indexes of a typical foundry enterprise in solving different scale simulation experiments In different time periods,the actual scheduling effect of intelligent algorithm aided scheduling mode and manual scheduling mode is compared.Experimental results show that the optimization rates of the four improved strategies are6.6%,22.9%,23.2% and 55.6% respectively,and the overall optimization rate is 72.8%.Multi algorithm experiments show that the improved genetic algorithm is superior to the other four algorithms in convergence speed,accuracy,speed and stability.The actual scheduling results show that,compared with the manual scheduling mode,the average scheduling time of the improved genetic algorithm aided scheduling mode is reduced from7 minutes to 2.5 seconds,and the average optimization rates of the standard deviation of the sum of grinding coefficients,the maximum completion time and the average matching coefficient are 93.5%,33.6% and 41.9% respectively,which effectively realizes the efficiency,stability and scheduling method of grinding scheduling The balance and matching of cases.